1、 2016 Her Majesty the Queen in Right of Canada, as represented by the Minister of Natural Resources 63ABSTRACTThis paper addresses the challenge of improving theperformance of heat pumps (HPs) in cold-climate conditionsby applying refrigerant mixtures. The potential benefits ofimplementing R-32/CO2z
2、eotropic refrigerant mixtures inthreedifferentresidentialair-sourceHPsforcoldclimatesarestudied.ThecasesconsideredareconventionalresidentialHP,HPwithavariablemixturecontrolsystem,andHPwithavari-ablecompressorspeed.Theseasonalperformanceofaheatingsystemwiththeseair-sourceHPs,supplementedwithanauxil-i
3、ary electric heater, is studied in the cold-climate city ofMontreal,Canada.Tothisaim,adetailedscreeningHPmodelpreviously developed is modified and used. The obtainedresultshighlightthepotentialHPperformanceimprovementofapplying refrigerant mixtures.INTRODUCTIONHeat pumps (HPs) have attracted a great
4、 deal of attentionas one of the most adapted heating solutions for meeting lowenergy consumption requirements in buildings. However,improving the performance of the HPs at low ambienttemperatures is still an open challenge. When there is a hightemperature difference between the cold source and the h
5、eatsink,thecapacityandthecoefficientofperformance(COP)ofthe HP fall drastically. This problem is the main reason for thelimited application of air-source HPs, especially in cold-climate regions such as are found in Canada.HP performance has remarkably improved over recentyears because of development
6、s made in variable-speedcompressors, control strategies, compact heat exchangers,lubrication, multistaging, and refrigerant injection. Signifi-cantimprovementsweremadeinthiswayonthesystemratherthan on components, but they are often complex. Advantagesanddisadvantagesoftheseimprovementsaresummarizeda
7、ndcomparedinpreviousresearch(Bertschetal.2005;Chuaetal.2010). Despite all these improvements, improving HP perfor-mance in cold climates and their environmental impacts hasbeen an ongoing concern.The new generation of air-source HPs has improvedperformance in cold climates through multistaging orove
8、sized compressors (Bertsch and Groll 2008). RefrigerantvaporinjectionwasalsoinvestigatedexperimentallybyWanget al. (2008) for residential HP systems operating on R-410Aand obtained a 30% capacity improvement at 17.8C (0F).Furthermore, in cold climates, applying hybrid heatingsystems (electric + HP)
9、is necessary to satisfy the indoorcomfortrequirements.Thesesolutionsincreasetheinitialcostof the system. Alternatively, geothermal HPs can be used toovercome poor cold-climate performance, but the high instal-lation cost hinders their widespread adoption.One of the promising low-cost solutions to ma
10、king HPsreasonablyefficientattemperaturesdownto30C(22F)orbelow is the use of refrigerant mixtures. Previous research hasreported an increase in HP performance when using zeotropicrefrigerant mixtures (Comakli et al. 2009; Hakkaki-Fard et al.2014a). The use of refrigerant mixtures with the aim ofincr
11、easing the COP and the heating capacity at low tempera-tures is an area that has not been widely researched to date(Hakkaki-Fard et al. 2014b).This paper is a continuation of the previous work of theauthorstoassessthepotentialofapplyingrefrigerantmixturesinresidentialHPsincold-climateconditions.Inpr
12、eviousstud-ies by the authors (Hakkaki-Fard et al. 2014a, 2014b, 2014c,2015), it was shown that the mixture of R-32 and carbon diox-Evaluation of Refrigerant Mixtures inThree Different Cold-ClimateResidential Air-Source Heat PumpsA. Hakkaki-Fard, PhD Z. Aidoun, PhD P.Eslami-Nejad,PhDA. Hakkaki-Fard
13、is an assistant professor at the Center of Excellence in Energy Conversion, School of Mechanical Engineering, SharifUniversity of Technology, Tehran, Iran. Z. Aidoun and P. Eslami-Nejad are researchers at Natural Resources Canada, CanmetENERGY,Varennes, QC, Canada.ST-16-006Published in ASHRAE Transa
14、ctions, Volume 122, Part 2 64 ASHRAE Transactionside (R-32/CO2) had the best performance among the refriger-ant mixtures considered. The current study assesses theseasonal performance of three different HPs with the R-32/CO2refrigerantmixture(conventionalresidentialHP,HPwithvariable mixture control
15、system, and HP with variablecompressor speed) in the cold-climate city of Montreal,Canada. In this study, R-410A is used as a reference refriger-ant because it is currently widely used in HPs.SYSTEM CONFIGURATIONFigure 1 shows schematics of (a) a conventional HP and(b) a HP with a variable mixture c
16、ontrol system. A HP with avariable-speed compressor has the same configuration as aconventionalHP.Theonlydifferenceisthatthevariable-speedcompressor takes advantage of a variable-speed drive tocontrol the speed of the compressor. The drive will increasethe compressor speed to compress more refrigera
17、nt whenhigher load is required and vice versa when lower load isrequired.A HP with a variable mixture control system consists ofa conventional HP composed of a condenser, an evaporator, acompressor, and an expansion valve and is equipped with asimple variable mixture control set up. The latter is in
18、tegratedbetween the evaporator and the compressor and is composedof two accumulators, a heater, and a rectifying column. WhenanincreaseinthecapacityoftheHPisrequired,theexpansionvalve opens and the accumulator above the rectifying columnispartiallyfilledwithliquidrefrigerant.Thisrefrigerantflowsto t
19、he low accumulator and enriches the lower boiling compo-nentflowingtothecompressorandhence,thesystemcapacityincreases. When a decrease in capacity is required, the lowaccumulatorisheatedandthehigherboilingcomponentinthecycle is enriched. More information on this process and otherpotential controllin
20、g systems can be found in the research byHalm et al. (1999).HP Model Mathematical BasisIn this work, only the main features and modificationsmade to the mathematical model previously developed byHakkaki-Fard et al. (2014a; 2015) are outlined. Some generalassumptions to develop the theoretical models
21、 are as follows:HP is used for heating only (for heating-dominated cli-mates)All system components are operating under steady-stateconditionsPressure drop in connecting tubes is neglectedFlow for the refrigerant inside tubes is one-dimensionalHeat loss to the surroundings is assumed to be negligible
22、Heat required to defrost the evaporator is not consideredThe expansion valve undergoes an isenthalpic processSaturated liquid and saturated vapor conditions areassumed at the condenser outlet and compressor inlet,respectivelyA scroll compressor is considered in the simulations. Therefrigerant mass f
23、low rate is calculated by the followingrelation:(1)In this study, the compression process is assumed to bepolytropicwithconstantefficiency.Thevolumetricefficiency,v,isextractedfromthecommercialcompressorperformancecurves.Cross-flow heat exchange is assumed in both thecondenser and the air-source eva
24、porator. The air-source heatexchangers are assumed to be finned tube. The overall heattransfer coefficient UiAiis obtained by the following relation:(2)(a)(b)Figure 1 Schematics of (a) conventional HP system and (b)HP with variable mixture control.mrmrvVcpsuc=UiAi1Ri-1RrRwRair+-11hrAr-ro/ riln2kL-10
25、hairAair-+-= =Published in ASHRAE Transactions, Volume 122, Part 2 ASHRAE Transactions 65To find the overall heat transfer coefficient using Equa-tion2,thefollowingcorrelationsandcoefficientsarerequired:air-side heat transfer coefficient, one-phase refrigerant heattransfer coefficients, and condensa
26、tion and boiling correla-tions for refrigerant phase change. These coefficients areobtained using the correlations available in the literature(Hakkaki-Fard et al. 2014a). The pressure drop results fromthe sum of acceleration and frictional components of thelosses. It is calculated by means of a corr
27、elation developed byRohsenow et al. (1999).Each system component is calculated separately. Heatexchangers are divided into a number of control volumeelements. Fundamental conservation equations of mass,momentum,andenergy,basedontheappropriatecorrelations,are applied to each control volume element. T
28、he system ofequationsobtainedisnonlinearanditsparametersarestronglylinked.Aniterativemethodisthereforeappliedtosolvethesetof equations. A gradient-based optimization method is used toupdate the iterations to speed up the convergence of theprogram. A separate subroutine is developed for each compo-ne
29、nt.The code is developed in FORTRAN, to whichREFPROP Version 9.1 subroutines (Lemmon et al. 2013) arelinked, to calculate the thermodynamic properties of air andthe refrigerants. The calculation steps are presented in thelogic flow chart (Figure 2).TheperformanceoftheHPisexpressedbyitsCOP,whichisthe
30、ratiooftheheatdeliveredtotheoverallworkofthecycle.It is defined as follows:(3)SIMULATIONTo compare the three HP systems with refrigerantmixtures,theirseasonalperformancesareassessedforthecityofMontreal(ClimateZone5).Figure3apresentsthetempera-ture distribution data taken from the National Solar Radi
31、ationDatabase typical meteorological year (TMY) 2 weather file(SEL 2008).The building loads used for the analysis were obtainedusing an energy model for a 210 m2(2260 ft2) single-familydetached home located in Montreal using the TRNSYSversion17 simulation tool (SEL 2013). The housing modeldeveloped
32、for this analysis was based on a common 1980sCanadian house modeled using the multizone building compo-nent (Type 56a). Information on the energy model validationcan be found in the paper by Kegel et al. (2012). The heatingloads obtained are presented in Figure 3b. The peak hourlybuilding load is 14
33、.29 kW (48760 Btu/h) with annual spaceheating requirements of 29,266 kWh (99.86E+06 Btu). Therelatively small load in summer is caused by the space-heatingrequirements of the basement.MODELING RESOLUTIONThe objective of this study is to assess the performance ofthethreeHPoptionsselectedforsatisfying
34、identicalspace-heat-ing requirements of a residential house for which cooling andhot-water loads are ignored at this stage. In this case, HP oper-ation,supplementedbyanelectricheaterasabackup,isconsid-ered.Thesystemenergyconsumptionisdefinedhereasthesumof the HP energy consumption (compressor and fa
35、ns) plus theenergy consumption of the backup electric heater.The system simulation was performed using a code devel-opedinFORTRANforthispurpose(Hakkaki-Fardetal.2015).The logic flow chart of the heating system is shown in Figure 4.System operation is controlled by using the heat load requiredto main
36、tain the building comfort condition. The backup electricheater will provide heat only if the performance of the HP is notsufficient to cover the hourly heating demand.CASE STUDIESSeasonal performances were simulated according to thesystem described in the previous section. Simulations wereperformed
37、for a total of five different cases: the three differentHPsusingtheR-32/CO2mixtureasrefrigerant,aconventionalHP, and another with a variable-speed drive for the compres-sor, both using refrigerant R-410A. The R-32/CO2(80/20)mixture was chosen as a constant mixture because it wasshown to have the hig
38、hest performance among other compo-sitions. It was assumed that all HPs have the same evaporator,Figure 2 Logic flow chart of the HP cycle.COPQHWCOMPWFEVAPWFCND+-=Published in ASHRAE Transactions, Volume 122, Part 2 66 ASHRAE Transactionscondenser, and compressor. The HP with variable mixturecomposi
39、tionwasequippedwithcomposition-changingequip-ment. HP specifications used for the simulations are given inTable1.Theheatexchangertubesandfinswereassumedtobemade of copper and aluminum, respectively. The compressorspeed was assumed to be 50 rps for constant-speed compres-sors and to vary from 30 to 1
40、00 rps for the one with a drive.Results and DiscussionFigures 5 and 6 present the heating capacities and COPsof different HPs as a function of outdoor air temperature.Table 2 presents the building load, the HP contribution, theelectric heater contribution, the system energy consumption,the energy co
41、nsumption reduction in comparison to the R-410Asystem,andtheseasonalsystemCOP.Moreover,Figure7 presents the HP and the electric heater contributions forhome heating needs. It can be seen that the R-32/CO2(80/20)mixture in the case with variable speed shows the best perfor-mance and its energy consum
42、ption is 27.38% lower incomparison to the conventional HP with R-410A. In addition,it has the best seasonal COP of 2.98. It is also shown that thebackup electric heater contribution in this case is only 2.4% ofthe building load.R-32/CO2(80/20) with a drive system is followed by R-32/CO2with a variab
43、le-composition system. Its energyconsumptionis23.10%lowerincomparisontoaconventionalHP with R-410A.(a)(b)Figure 3 (a) Weather temperature data and (b) building heating load for the simulated home in Montreal.Figure 4 Logic flow chart of the heating system.Published in ASHRAE Transactions, Volume 122
44、, Part 2 ASHRAE Transactions 67It can be seen further that by using a variable compressorspeed the energy consumption of the system with R-410A canbe reduced by 20.84%. However, the lowest-cost solution thatdoes not add any extra component to the system is replacingR-410A with the R-32/CO2(80/20) mi
45、xture; it can reduce thesystem energy consumption by 12.15%.Figures 812 present the seasonal cumulative buildingload, HP energy consumption, backup electric heater energyconsumption, and system energy consumption of the heatingsystemfortheserefrigerantsasafunctionoftimeforthesimu-lated house model i
46、n Montreal. According to these results R-32/CO2(80/20) with drive system shows the best annualperformance, followed by R-32/CO2with variable composi-tion system.CONCLUSIONIn this study, an assessment of R-32/CO2refrigerantmixture in three different residential HPs (conventional resi-dential HP, HP w
47、ith variable mixture control system, and HPwith variable compressor speed) is performed. Seasonalperformance of these HPs is assessed using a detailed numer-ical simulations and a comparison with a reference systemusing refrigerant R-410A is made. Simulations are performedfor the cold-climate city o
48、f Montreal. The obtained resultsshow that the HP with variable compressor speed using R-32/CO2(80/20)cansaveupto27%incomparisonwithaconven-tional HP using R-410A as a refrigerant. The HP with variableR-32/CO2composition can save up to 23%. Moreover, by justreplacing R-410A with R-32/CO2(80/20) without anychanges to the s
